Fix dynamic contempt for MultiPV

Use rootMoves[PVIdx].previousScore instead of bestValue for
dynamic contempt. This is equivalent for MultiPV=1 (bench remained the
same, even for higher depths), but more correct for MultiPV.

STC (MultiPV=3):
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 2657 W: 1079 L: 898 D: 680
http://tests.stockfishchess.org/tests/view/5aaa47cb0ebc590297330403

LTC (MultiPV=3):
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 2390 W: 874 L: 706 D: 810
http://tests.stockfishchess.org/tests/view/5aaa593a0ebc59029733040b

VLTC 240+2.4 (MultiPV=3):
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 2399 W: 861 L: 694 D: 844
http://tests.stockfishchess.org/tests/view/5aaf983e0ebc5902a182131f

LTC (MultiPV=4, Skill Level=17):
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 747 W: 333 L: 175 D: 239
http://tests.stockfishchess.org/tests/view/5aabccee0ebc5902997ff006

Note: although the ELO differences seem huge, they are inflated by the
nature of Skill Level / MultiPV search, so I don't think they can be
reasonably compared with classic ELO strength.

See https://github.com/official-stockfish/Stockfish/pull/1491 for some
verifications searches with MultiPV = 10 at depths 12 and 24 from the
starting position and the position after 1.e4, comparing the outputs
of the full PV by the old master and by this patch.

No functional change for MultiPV=1

diff --git a/src/search.cpp b/src/search.cpp
index 1a516ff80b1e7a1c7a95d9d1ac9c9cfd8412e167..7b7737ea9fd2a946303a66b0af53d8235a30a766 100644 (file)
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -343,14 +343,15 @@ void Thread::search() {
           // Reset aspiration window starting size
           if (rootDepth >= 5 * ONE_PLY)
           {
           // Reset aspiration window starting size
           if (rootDepth >= 5 * ONE_PLY)
           {

+              Value previousScore = rootMoves[PVIdx].previousScore;

               delta = Value(18);
               delta = Value(18);

-              alpha = std::max(rootMoves[PVIdx].previousScore - delta,-VALUE_INFINITE);
-              beta  = std::min(rootMoves[PVIdx].previousScore + delta, VALUE_INFINITE);
+              alpha = std::max(previousScore - delta,-VALUE_INFINITE);
+              beta  = std::min(previousScore + delta, VALUE_INFINITE);

 
               ct =  Options["Contempt"] * PawnValueEg / 100; // From centipawns
 
 
               ct =  Options["Contempt"] * PawnValueEg / 100; // From centipawns
 

-              // Adjust contempt based on current bestValue (dynamic contempt)
-              ct += int(std::round(48 * atan(float(bestValue) / 128)));
+              // Adjust contempt based on root move's previousScore (dynamic contempt)
+              ct += int(std::round(48 * atan(float(previousScore) / 128)));

 
               Eval::Contempt = (us == WHITE ?  make_score(ct, ct / 2)
                                             : -make_score(ct, ct / 2));
 
               Eval::Contempt = (us == WHITE ?  make_score(ct, ct / 2)
                                             : -make_score(ct, ct / 2));